A half-scale model bridge deck has been tested to collapse in the TRRL Structures Laboratory. It consisted of eight precast prestressed beams with an in-situ concrete slab. The design conformed to BS 5400: Part 4 and contained no arbitrary reserves to strength. The deck was tested under dead load, superimposed dead load and live load as specified in BS 5400: Part 2. It was loaded to collapse by increasing the HB component of load until failure. Member forces calculated from a linear grillage analysis (GRIDS) are used to evaluate the test results. The report contains: an evaluation of reserves of strength in the design according to code calculations and the influence of the form of loading; a discussion of the effectiveness of code provisions; an evaluation of the reserves of strength present in the model and the influence of material strength, redistribution, etc; and an indication of how the reserves of strength might be used in the assessment of similar structures. It is concluded that the design which complied with BS 5400 was conservative, but the overall margin was not particularly large. The model failed in beam shear at 3.17 times the HB load factored for the ultimate limit state. The unfactored code calculation indicated failure at 2.8 times the HB load. In some areas, there were reserves of strength which could possibly be used in the assessment of similar structures. Examples of this are flexure in the slab, torsion in the beams and end diaphragm, interfacial shear, web crushing, combined shear and torsion and shear strength close to the support. In these cases, unfactored code calculations generally indicated failure at substantially less than 2 times the HB load. (A)
Abstract